As promised by Elon Musk, Twitter code related to Suggested tweets users in GitHub published and also explained about this decision by publishing a new post on his blog.
according to Twitter Description“Algorithm of suggested posts consists of three main stages”. It first collects “the best tweets from various recommended sources,” then ranks those tweets with a “machine learning model.” Finally, tweets from people you’ve blocked, tweets you’ve already seen, or tweets that aren’t safe will filter out before putting them on your timeline.
How to filter Twitter’s suggested tweets
This post also explains each step of the process in more detail. For example, it notes that around 1,500 tweets will be processed in the first phase, and the goal is to display about 50% of tweets from people you follow and 50% from unfollowers in your For You timeline.
Additionally, it explains that the ranking is designed to “optimize positive engagement (likes, retweets, and replies)” and ultimately tries not to show you too many tweets from the same person.
More details related to this code become clear when using it, which some researchers including “Jin Manchun Wong” They are using it now.
Elon Musk, the CEO of Twitter, has mentioned such a decision for a long time. In March of last year, before becoming the company’s new owner, he asked his followers for their opinion on open-sourcing Twitter’s algorithm, and about 83 percent agreed with the decision. He then said last month that the code for Twitter’s suggested tweets would be open-sourced next month, which he has now delivered on his promise and can be accessed via GitHub.
